self-generating-skills

To convert repeated AI workflows into reusable skills, draft a new SKILL.md file on demand using a skill-drafting command that translates your task description into a standardized Pi skill directory entrypoint.

Skill authoring requires YAML frontmatter with name and description, a clear usage description, triggers and non-triggers, verification steps, and safety guidance for destructive actions to ensure proper workflow templating.

Generating a new skill does not require external dependencies or components, as the self-generating-skills process drafts the SKILL.md file entirely on demand within your existing project structure.

For deduplication when standardizing recurring tasks, fold in learned edge cases after successful reuse, remove non-generalizable context, and check for overlap with existing skills using the curator workflow.

Workflow templating for authoring recurring tasks works by applying an authoring checklist that ensures each generated skill includes a precise description, clear scope, required tools, and safety notes before deployment.

You should not use automation to generate a new skill directory for workflows that are highly specific, contain non-generalizable context, or lack clear verification steps and safety boundaries for destructive actions.